Exploring Temporal Context at Multiple Scales for Crop Mapping with Fully Convolutional Recurrent Nets and Fully Connected CRFS.
Marcos RogozinskiJorge Andres Chamorro MartinezPatrick Nigri HappRaul Queiroz FeitosaPublished in: IGARSS (2021)
Keyphrases
- fully connected
- multiple scales
- temporal context
- multiscale
- conditional random fields
- spatial context
- scale space
- temporal information
- scale free
- spatio temporal
- audio visual
- pixel wise
- feed forward
- recurrent neural networks
- image analysis
- machine learning
- complex networks
- spatial and temporal
- maximum likelihood
- markov random field
- higher order
- information extraction
- image processing